The rapid explosion of information and data on biologically important molecules requires powerful computer software to aid scientists with integrating and understanding that information. As the number, size, and types of on-line information and databases increase, it becomes unreasonable to assume that working scientists will have a full understanding of the organization and contents of these information resources. The hypothesis of our work in the design and development of an intelligent assistant for molecular biologists is that the technology of knowledge-based systems is sufficiently mature to assist in the discovery and interpretation of information relevant to molecular biological research. Our long-term goal is to increase the productivity of biomedical scientists by designing, implementing, and providing computer-based intelligent assistants for them. There are limitless numbers of such assistants that can be built, some of which already exist in some form, ranging from assistants that manage routine activities intelligently (e.g., data collection) to those that assist in creative tasks (such as explaining anomalies and planning experiments). Creating a uniform, easy to use environment for this broad collection of assistants becomes an essential part of the goal since we can only save scientists' time if they can use the software we provide. Thus we envision a biochemist's workstation with which scientists can exercise a wide variety of computer-based tools, including specialized assistants, matching algorithms, simulations, and databases. Specifically, we propose to design and develop a workstation which will function as an Intelligent Biomedical Assistant (IBA) for researchers in molecular biology and genetics. This IBA will be a highly adaptive and extensible system which will employ techniques from Artificial Intelligence (particularly expert systems and concept learning) which will assist the researcher in model building, experiment planning and hypothesis testing. The IBA will be assembled to assist with structural and functional questions about proteins that bind to DNA. It will have knowledge of some of the existing software and databases that currently exist to help researchers, and will include basic knowledge of molecular biology in which questions are framed.